Background on Review Contributor

Toni Seiler BSpTher, MS SpPath
Director of ELR Software Pty Ltd

Developer of eLr (Extra Language Resource) software, and provider of consultancy services in advice, design and development of computerised solutions in fields relating to speech, language and literacy impairments.

What does SpeechViewer III Do?

SpeechViewer III is a computer program which provides visual and auditory feedback in response to sound. It is designed to be used within a therapy session where the clinician’s goal is to improve the speech production of the client. The program presents the clinician with a range of “exercises” which target different aspects of speech production. These range from very fundamental goals (like making any sound), to very specific goals (like producing a specific sound or phoneme, or producing voice at a particular pitch).

The first exercise aims to reinforce the client for producing sound or voice. In this exercise the graphic on the screen responds whenever the client produces sound. There is a choice of 4 graphics – suitable for young children or adults.

The next exercise enables the client to see when they have produced a sound or voice which is loud enough. In this task, the clinician records a sound (the target), where the loudness level is represented on the screen (eg by a balloon getting bigger). The client then tries to match the loudness level of the clinician’s model. Again, the graphics are suitable for paediatric or adult populations. There are similar tasks designed to allow the client to see when their speech attempt has matched the clinicians in pitch range, or pitch control.

The phonology exercises allow the client to get visual feedback when their attempt to produce a particular sound matches the recorded model. The first exercise focuses on the production of a single phoneme (or sound). When an accurate production occurs, something happens on the screen (eg a man picks an orange from a tree).

The next phonology exercise requires with client to produce two different sounds. The feedback is like a game, eg a car in a maze. When one sound is produced the car goes to the left, and when the contrasting sound is produced the car goes the other way. The goal is to miss obstacles in the way of the car. This is particularly powerful, as the client can clearly see which sound s/he is producing. The software is sensitive enough that it can differentiate between very similar sounds (eg “f” and “s”, or “s” and “sh”). There is a further exercise which allows for production of 4 sounds, and the game responds depending on which sound is produced.

Another phonology exercise encourages blending of two to four sounds. For example, “c” “ah” “t”. For example, the client sees the boy at the bottom of an incline. By producing the sounds one after the other, the boy climbs the incline, and is therefore encouraged to say “cart”, by blending the 3 sounds.

Who would this be suitable for?

1. Clients who are unable to produce or repeat sounds:
   • Hearing impaired
   
   
   • Children who have not understood the power of speech, i.e. that when you make a noise something happens (stimulus response). Many autistic children are unable to repeat sounds, and have not made the connection between speech and communication. In this game, they can see and hear the result of making sounds.
   
   
   • Children who understand communication, but are unable to produce the sounds they want (dyspraxia). This program provides specific reinforcement within the context of a game, and is powerful enough to respond only when a particular sound is produced.
   
   
   • Adults with acquired speech and language disorders, eg following stroke or head injury. These clients may also be dyspraxic, and have trouble voluntarily coordinating the appropriate muscle movements to produce sounds, or may have muscular weakness and produce inaccurate sounds.


2. Clients needing to modify voice quality.
   • Adults and children with voice disorders who need to work on loudness levels and pitch range.
   
   
   • Accent modification

• It can accurately discriminate between very similar sounds. The clinician can also adjust the setting so that the reinforcement occurs when there is an approximation of the sound. As the client improves the settings can then be modified so that the client needs to be more accurate.


• The reinforcement occurs instantly – excellent for behavioural reinforcement, particularly with young autistic children or those with dyspraxia


• Because the activities are within the context of a game, it may even make it easier for dyspraxic clients to produce the sounds, as it could become less of a “voluntary” movement and more automatic. It would be interesting to observe this.


• The manual is good. It give clear descriptions of each of the activities, the goals, and how to adjust settings.


• It enables the clinician to record each of the client’s sessions, so that comparisons over time can be made.


• It allows for the records to be imported into other documents, such as reports.


• It provides for detailed analysis of client responses by displaying speech samples via spectrograms and frequency / pitch 3 dimensional graphs. The spectrogram can be “frozen” after each trial, so that the client can keep trying until their responses more closely matches the target model. All of these trials can be saved so that comparison over time can occur.


• The graphics are clever. There appropriate choices for adults and paediatric. The child oriented graphics work on different levels. If the client is cognitively able to understand the game, it can work as a game. If they are unable to understand the concept of the game (eg the man picking oranges), they will still receive reinforcement by the movement changes on the screen.


• The games are fun.

Technical comments

The version reviewed is apparently able to work with Windows 3.2, OS2 as well as current versions of Windows. Perhaps this is the reason for its somewhat antiquated interface. From the initial installation process which does not allow the usual choice of destination folders, to the programs insistence on changing the screen resolution to 256 colours, to the file-save/load menus, this program looks like an “early Windows” program. It also relies heavily on drop-down menus for controls and selections.

A more modern interface would certainly include “toolbar” buttons for more simple, intuitive controls. However, in its favour, it does actually have “speed keys” for almost all tasks (eg the PageUp/Down keys can be used to toggle between client and clinician inputs). It was also stable and fast on the test machines which were early model Pentium 200-300 machines. The first had an “on-the mother board” Sound Blaster compatible sound card, while the second had a “stand-alone” SB card. Although the first machine allowed many tasks to work, the quality of the recording did not allow useable preparation of sample phonemes.